Re: a lot of shared buffers hit when planning for a simple query with primary access path

Pavel Stehule <pavel.stehule@gmail.com>

From: Pavel Stehule <pavel.stehule@gmail.com>
To: David Rowley <dgrowleyml@gmail.com>
Cc: James Pang <jamespang886@gmail.com>, pgsql-performance@lists.postgresql.org
Date: 2024-07-01T10:20:15Z
Lists: pgsql-performance
Hi

po 1. 7. 2024 v 12:10 odesílatel David Rowley <dgrowleyml@gmail.com> napsal:

> On Mon, 1 Jul 2024 at 21:45, James Pang <jamespang886@gmail.com> wrote:
> >    Buffers: shared hit=110246         <<< here planning need access a
> lot of buffers
> >  Planning Time: 81.850 ms
> >  Execution Time: 0.034 ms
> >
> >    could you help why planning need a lot of shared buffers access ?
>
> Perhaps you have lots of bloat in your system catalogue tables. That
> could happen if you make heavy use of temporary tables. There are many
> other reasons too. It's maybe worth doing some vacuum work on the
> catalogue tables.
>

The planners get min/max range from indexes. So some user's indexes can be
bloated too with similar effect

Regards

Pavel


> David
>
>
>